Quantitative genetic studies have consistently demonstrated a heritable component for specific language impairment (SLI). However, only a handful of groups have begun assessing the genetic epidemiology of SLI. In a previous phase of our ongoing research, we found compelling evidence for a risk polymorphism within 13q21-22, which was subsequently replicated. However, the specific susceptibility allele that acts to increase risk for SLI within 13q21 region has not yet been identified. Equally important, previous studies have not examined the genetic etiology or etiologies of SLI in the context of the multiple, quantitatively distributed underlying language processes that may lead to SLI. Thus, the relationship between the substantial clinical heterogeneity among individuals with SLI and concomitant genetic heterogeneity remains largely unexplored. The goal of this application, therefore, is to address three specific aims. We proposed to localize SLI susceptibility alleles by applying a combination of multi-level approaches. We will use bioinformatics and molecular approaches to identify possible susceptibility-allele-harboring sequences within 13q21 (Aim 1). We will extend our family collection to increase power to detect novel SLI loci and SLI-related QTLs (Aim 2). We will employ new multivariate approaches which are expected to better refine our localization, increase power, and allow us to examine these loci in the multivariate context of several underlying language processes from the SLI literature (Aim 3). These aims are important because they link genetic analysis with the multiple cognitive pathways that may lead to SLI in the hopes of better identification and treatment.